ImageNet Classification using WordNet Hierarchy
ImageNet Classification using WordNet Hierarchy
Embedded Files
Convolutional neural networks (ConvNets) have become increasingly popular for image classification tasks. All contemporary computer vision problems are being dominated by ConvNets. Conventional training methods using cross-entropy loss for training have constantly outperformed the state-of-the-art technique to set a new standard in the ImageNet classification challenge. However, growing accuracy come at the cost of enormous number of parameters and computations. Further, classical learning algorithms do not utilize the semantic relationship between the classes present in the dataset. Thus, interpreting the behavior of the model become difficult even though the results may be desirable. Hence, we demonstrate a classification method by leveraging the WordNet hierarchy on the ImageNet dataset to establish class relationships and label embedding. The model is trained using cross entropy with soft labels based on the semantic similarity between the generated output and the ground truth. Unlike categorical cross entropy, it does not treat every predicted label as equally erroneous. The method generates meaningful neighboring classes in the feature space of the true label.
Convolutional neural networks (ConvNets) have become increasingly popular for image classification tasks. All contemporary computer vision problems are being dominated by ConvNets. Conventional training methods using cross-entropy loss for training have constantly outperformed the state-of-the-art technique to set a new standard in the ImageNet classification challenge. However, growing accuracy come at the cost of enormous number of parameters and computations. Further, classical learning algorithms do not utilize the semantic relationship between the classes present in the dataset. Thus, interpreting the behavior of the model become difficult even though the results may be desirable. Hence, we demonstrate a classification method by leveraging the WordNet hierarchy on the ImageNet dataset to establish class relationships and label embedding. The model is trained using cross entropy with soft labels based on the semantic similarity between the generated output and the ground truth. Unlike categorical cross entropy, it does not treat every predicted label as equally erroneous. The method generates meaningful neighboring classes in the feature space of the true label.
Paper: https://ieeexplore.ieee.org/abstract/document/10189102
Paper: https://ieeexplore.ieee.org/abstract/document/10189102
Clustering with Multi-Layered Perceptron
Clustering with Multi-Layered Perceptron
Multi-layered perceptron (MLP) is a widely used neural network architecture for supervised learning. The feed-forward network maps unknown data to a label based on prior labeled training samples. The accuracy with which test data is classified correctly depends on the number of training samples. However, in unsupervised learning hidden grouping patterns in data are learnt through various exploratory analysis without any prior knowledge of labels. Clustering using MLP architecture requires defining a different type of loss function. In this work, we theoretically analyse the effectiveness of a loss function previously reported for this purpose, where the input features are mapped to an output cluster node based on the degree of belongingness. Thus, similar samples are grouped together to form clusters using entropy based measures. A detailed study on the nature of the loss function and its convergence properties have been presented in two theorems.
Multi-layered perceptron (MLP) is a widely used neural network architecture for supervised learning. The feed-forward network maps unknown data to a label based on prior labeled training samples. The accuracy with which test data is classified correctly depends on the number of training samples. However, in unsupervised learning hidden grouping patterns in data are learnt through various exploratory analysis without any prior knowledge of labels. Clustering using MLP architecture requires defining a different type of loss function. In this work, we theoretically analyse the effectiveness of a loss function previously reported for this purpose, where the input features are mapped to an output cluster node based on the degree of belongingness. Thus, similar samples are grouped together to form clusters using entropy based measures. A detailed study on the nature of the loss function and its convergence properties have been presented in two theorems.
Taxonomy Driven Learning of Semantic Hierarchy of Classes
Taxonomy Driven Learning of Semantic Hierarchy of Classes
Standard pre-trained convolutional neural networks are deployed on different task-specific limited class applications. These applications require classifying images of a much smaller subset of classes than that of the original large domain dataset on which the network is pre-trained. Therefore, a computationally inefficient and over-represented network is obtained. Hierarchically Self Decomposing CNN addresses this issue by dissecting the network into sub-networks in an automated hierarchical fashion such that each sub-network is useful for classifying images of closely related classes.
Standard pre-trained convolutional neural networks are deployed on different task-specific limited class applications. These applications require classifying images of a much smaller subset of classes than that of the original large domain dataset on which the network is pre-trained. Therefore, a computationally inefficient and over-represented network is obtained. Hierarchically Self Decomposing CNN addresses this issue by dissecting the network into sub-networks in an automated hierarchical fashion such that each sub-network is useful for classifying images of closely related classes.
Page updated
Google Sites
Report abuse